Optimal lower estimates for eigenvalue ratios of Schrödinger operators and vibrating strings (Q2568561)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 2215481
Language Label Description Also known as
default for all languages
No label defined
    English
    Optimal lower estimates for eigenvalue ratios of Schrödinger operators and vibrating strings
    scientific article; zbMATH DE number 2215481

      Statements

      Optimal lower estimates for eigenvalue ratios of Schrödinger operators and vibrating strings (English)
      0 references
      0 references
      18 October 2005
      0 references
      Considered is the eigenvalue problem \(-y''+q(x)y=\lambda y\) with \(q\leq 0\) and Dirichlet boundary conditions and all eigenvalues positive. It is shown that the eigenvalues satisfy \[ \frac{\lambda _m}{\lambda _n}\geq \left(\left[ \frac mn \right]\right)^2 \] for \(m>n\geq1\), with equality if and only if \(m\) is a multiple of \(n\). A similar result holds for Neumann boundary conditions. This complements estimates on eigenvalue ratios for the case \(q\geq0\) obtained by \textit{M. S. Ashbaugh} and \textit{R. D. Benguria} [Commun. Math. Phys. 124, No. 3, 403--415 (1989; Zbl 0677.34027)] and \textit{Y.-L. Huang} and \textit{C. K. Law} [Proc. Am. Math. Soc. 124, No. 5, 1427--1436 (1996; Zbl 0966.34024)]. Using the Liouville transform, the above result is applied to the vibrating string eigenvalue problem \[ -y''=\mu\rho(x)y \] on \((0,1)\) with Dirichlet boundary conditions, where \(\rho\) is concave and positive. It is shown that the eigenvalue ratios satisfy \[ \frac{\mu_m}{\mu_n}\geq \left(\left[\frac mn\right]\right)^2 \] for \(m>n>1\), and if \(\rho\in C^2([0,1])\), that equality holds if and only if \(\rho\) is constant and \(m\) a multiple of \(n\). \textbf{Note:} Statement and proof of Lemma 4.1 are incorrect. The function \(\tilde \rho_\epsilon \) belongs to \(C^1([0,1])\), being \(C^\infty \) on the subintervals \([0,\epsilon ]\), \([\epsilon ,1-\epsilon ]\), \(1-\epsilon ,1]\). (The authors state that it is \(C^\infty \) on \([0,1]\).) \(C^1\) is needed in the application to ensure that its second derivative (in the weak sense) is \(L^1\). The function \(\tilde \rho_\epsilon \) given in the proof is not \(C^1\) at \(\epsilon \) and \(1-\epsilon \), in general, but it can easily be modified to become \(C^1\), e.\,g., putting \(L_1(x)=\rho_\epsilon (\epsilon )+\rho_\epsilon '(\epsilon )(x-\epsilon )\), still satisfying \(\tilde\rho _\epsilon \to\rho \) in \(L^1\).
      0 references
      Schrödinger operator
      0 references
      vibrating string equation
      0 references
      eigenvalue ratio
      0 references
      modified Prüfer transform
      0 references

      Identifiers